2. Team Details :
Member Name , Enrollment number, Gmail ID :
▪ TUSHAR SINGH , 01415611921, tusharsinghrawat.delhi@gmail.com
▪ JATIN SINGH SAGOI, 02115611921,
▪ ASHUTOSH DHAL ,01115611921,
3. Prologue :
▪ Work Flow for the Model
▪ Discussing the Problem
▪ Understanding the Approach
▪ Implementing the Solution
▪ Explanation for the Algorithms
▪ Scope of improvement
▪ The final Conclusion
4. Work Flow :
▪ Collecting and preprocessing of the
dataset
▪ Train and Test split
▪ Training Model using ANN
▪ Model Testing
▪ Prediction Using ANN
▪ Data visualization
▪ Model deployment Step 3
Prediction using
Model
Data visualization
Step 2
Train Test split Model training
Step 1
Data collection Data preprocessing
5. Discussing the Problem :
▪ Heart failure is a chronic and progressive condition where the heart is unable to pump enough
blood to meet the body's needs. This can occur due to various reasons, including damage to
the heart muscle from heart attacks, high blood pressure, heart valve problems, congenital
heart defects, or other underlying health conditions.
▪ Addressing the problem of heart failure requires a comprehensive approach involving
prevention, early detection, appropriate management, and patient education. Strategies
include promoting a healthy lifestyle, managing risk factors, providing access to quality
healthcare, optimizing heart failure treatments (such as medications, devices, and surgeries),
and advancing research to develop new therapies and interventions.
▪ By addressing the challenges associated with heart failure, we can work towards reducing its
impact and improving the outcomes for individuals affected by this condition.
▪ We are taking an initiative by introducing a heart failure prediction model to solve this critical
problem By using the ANN (Artificial Neural Network).
6. Approach :
▪ We are solving this critical problem by using the heart failure prediction
model based on ANN.
▪ We will be collecting the vast data to analyze the root cause for the heart
diseases and failures.
▪ After analyzing the cause , we will be designing a model which will predict
the heart failure based on the patient different entities like age, chest pain,
Anaemia , diabetes, High-Blood pressure and etc.
▪ By using the vast dataset we will be training our model and then analyzing
the patient data.
▪ Data visualization will be used for understanding and analyzing the reports.
7. Implementing the Approach :
▪ Taking a vast data set from websites like Kaggle , Data.gov and etc
for our Model.
▪ Importing the Dependencies like pandas, numpy, tensorflow,
seaborn and matplotlib for the functioning and data visualization
in our Model.
▪ Data collection and Pre-processing.
▪ Introducing the data-frames to convert the csv dataset file into a
tabular format for better understanding of the data.
▪ Using the Histogram for the data-points representation.
8. Solution :
▪ Splitting the data into Training data and Testing data in the 80 : 20
respectively.
▪ Training data is used to train the Model to produce the accurate
result for the raw input in the future.
▪ Model training using the ANN (Artificial neural network) which is a
computational model in Machine learning.
▪ Before deploying the Model , Model Evaluation is done on a
random input with the labelled data.
▪ Calculating the accuracy for both the training data and testing
data. If the accuracies are closed enough then it indicates toward
the balanced data-set.
9. Solution :
▪ Data Visualization using the seaborn and matplotlib
libraries
▪ Using different data-points like Smoking , age, diabetes, high blood
pressure to calculate the chances for the heart attack or failure.
▪ Representing the data is the most crucial part as it provides the
insight and helps in decision making process.
10. Explanation :
*Importing the dependencies:
Importing the required libraries for the functioning and data
visualization. Libraries like pandas , numpy , matplotlib , seaborn
and etc.
* Data collection & pre-processing:
This is use to remove any null, duplicate value from the data-set for
the proper functioning of the algorithm. Using a proper and large
dataset is necessary so that model training is well achieved and
accuracy is high as possible for the Model.
11. Explanation :
*Splitting Training data and Testing data:
We have to divide the dataset into two segments , one is training data which is
used to train the machine and create our model for the future inputs , and the
second is testing data which will be used to check the models accuracy.
Dataset must be balanced for high efficiency of the model.
*Building the predictive Model:
After the machine training, its time for the model evaluation by random input with
the labelled data.
12. Explanation :
*Data Visualization:
Libraries like seaborn, matplotlib and sklearn is used for the functionality like use of
histogram and graphs to represent the relationship between different entities(data-points) .
sklearn is a data analysis library used for data analysis and understanding.
*ANN (Artificial neural network) :
ANN stands for the Artificial Neural network which is a computational method used in the
machine learning . ANN structure is similar to the human brain . It consist of major 3 layers
: input layer, hidden layer and output layer. Each node in the ANN performs the simple
Computation. Once a ANN is trained it can be used for multiple task like regression,
classification and pattern recognition and etc.
13. Scope of Improvement :
• We have successfully achieved the high accuracy of 81% by using
the ANN .
• We can increases the accuracy by just introducing a large dataset .
• This model can also be design by other algorithms like KNN ,
Logistic Regression or Random forest.
14. The Conclusion :
▪ This model has the capability to predict the heart disease and
failures which can save lives and help to take the precautions
before any harmful disaster .
▪ This is just an initiative from our team to reduce the number of
heart attacks to save someone priceless life and help them to live
healthy.
▪ We are still improving our Model and trying to achieve the highest
accuracy as possible.